Peter L. Groves
Department of English,
Monash University, Melbourne
Email: Peter L. Groves
Copyright © Peter L. Groves 2005
Received: 5 March 2004; Published: 2005
ABSTRACT: The puzzle of Wyatt's metre was famously summed up by an early twentieth-century reviewer of a modern edition of his poems, who observed "At one moment he is the equal of the greatest in his command of rhythm and metre; at another he seems to be laboriously counting syllables on his fingers—and getting them wrong sometimes." This paper explores the origins of Wyatt's "proto-pentameter" in a misconstruction of the Italian endecasillabo, and tries to show that the mixture of familiarity and strangeness we find in his versification can be explained by the fact that he was writing as a pioneer at the very beginning of the modern English pentameter tradition. The strangeness arises from two causes: one is that he did not immediately see that in order to be intelligible as a metre, pentameter requires certain kinds of constraint on the metrical ordering of the line; the second is that he explored the expressive possibilities of certain prosodic features such as catalexis and the silent beat that have largely been avoided in the literary tradition of pentameter for essentially prescriptive reasons, though they have been exploited in Shakespeare's dramatic verse and in the more relaxed versification of certain C20 poets. It also contends that the radical inventiveness of Wyatt's contribution has been somewhat obscured by the traditional category of "accentual-syllabic" metre, which conflates merely ictosyllabic regulation, as in the case of poulter's measure, with what might be called positional or footed ("ictothetic") regulation.
KEYWORDS: Wyatt, metre, pentameter, accentual-syllabic, endecasillabo, ictothetic
At one moment he is the equal of the greatest in his command of rhythm and metre; at another he seems to be laboriously counting syllables on his fingers—and getting them wrong sometimes—and at a third he is, like some of his predecessors, floundering about for a foothold on stresses that may happen anywhere in the bog."[1]
The problem of Wyatt's metre is not hard to formulate: consider, for example, the well-known sonnet "The longe love" (example #2 below). What puzzles us is its apparently hybrid nature: six of its lines scan straightforwardly in the modern pentameter tradition that stretches from Spenser to Auden,[2] but another six don't even come close. If Wyatt (as many critics claim) wasn't attempting to write pentameters, or wasn't capable of it, why do so many of his lines succeed so spectacularly? If he was, why do so many fail? As the TLS reviewer quoted in the epigraph remarked, "It is more than an academic question. The doubt interferes with the reader's enjoyment of the poetry".
My investigation will be limited to Wyatt's circumdecasyllabic line, the line of the sonnets, the satires, and the Penitential Psalms, because that's where the peculiar problems seem to be (and in any case, the scansion of his songs is complicated in indeterminable ways by the influence of the lost music for which they were written [see Maynard]). I should add that since (as many commentators agree) his approach to metre seems to have been inventive and experimental, we perhaps should not expect that one system will necessarily account for every circumdecasyllable that he wrote. For example, as the reader can verify by experiment, rather fewer than one in three randomly chosen decasyllabic strings of English prose will be metrical when considered as modern iambic pentameter; given that there are decasyllabic poems of Wyatt, such as "Caesar, when that the traytour of Egipt," that seem to have just that proportion of metrical to unmetrical lines, such verse may represent experiments in a purely syllabic kind of metre.[3]
But the proportion of 'successful' lines in much of Wyatt's poetry is significantly higher than this, and is unlikely to have been produced by chance: there are poems, such as the sonnet "Dyuers dothe vse as I have hard and kno," that scan flawlessly as modern pentameter from beginning to end. The problem, then, is not that Wyatt wrote decasyllables that don't scan as conventional pentameters, but that he did not do so consistently. Of course, because metrical awareness of the pentameter is something that must be acquired, many readers will not notice the problem in the first place,[4] but the strange inconsistency remains, and while "the mystery of Wyatt will never be solved in a way that will satisfy all of his readers" (Woods 85), attempts that seek to explain only his failures should satisfy no-one.
It was sometimes assumed, for example, that he was just metrically inexperienced, or incompetent (perhaps a victim of what Eleanor Hammond called, in a desperate attempt to explain the puzzling metrics of the fifteenth century, "the cramping and inbreeding which weakened the intellectual fibre of the educated with every decade" [22]), but this is belied by the metrical skill that he frequently exhibits. A similar objection applies to the common but linguistically naïve claim that rapid changes in the language left Wyatt somehow "unsure about pronunciation and accent":[5] Wyatt's native competence in such matters is all too decisively vindicated by the wooden regularity of his verse in poulter's measure.
Others have seen him as attempting an undemanding metrical code that scarcely admits the possibility of failure, setting the bar so low that just about any circumdecasyllabic line would scrape over it. But if all Wyatt was aiming at was a loose four-to-six-beat verse with medial caesura (Lewis, Schwartz)[6] or a mere pair of rhythmical phrases, a sort of jointed free verse (Harding, Southall, Robinson), then it is hard to explain why his lines so frequently have ten syllables: according to Rebholz (49-50), 91.6% of the "long lines" in the autograph Egerton MS (E) are decasyllabic, and Daalder has shown that Wyatt's spelling of the Pentitential Psalms in E regularly varies orthography (e.g. <provokt> vs <provokyd> to maintain decasyllabicity. It is for this reason also that we must reject E. K. Chambers' more charitable notion that the sonnets are translator's notes waiting for "subsequent polishing" (122): there is no reason why such notes should have coincidentally fallen into decasyllables.
Moreover, the puzzle is not just that he stumbled across the odd pentameter: at times he seems to handle the line with extraordinary skill. As Kenneth Muir has said, "The success of 'They fle from me' depends [in part] on its rhythmical subtlety" (Life 245). This tends to suggest that whatever metre Wyatt was using in his most successful circumdecasyllabic poems was actual rather than notional. An actual metre is one in which unmetricality can be perceived directly, like the modern pentameter (if a student quotes a line of Pope as "What other planets attend other suns" we immediately know it must be a misquotation even if we don't know what the original is); a notional metre is one where unmetricality can only be discovered indirectly by inspecting the line (counting the syllables in a line of English "isosyllabic" verse, for example). Notional metres impose constraints in the way that actual metres do, but they do not produce intelligible metrical form. They may generate rhythmical effects of a rather general kind (no doubt Marianne Moore's dry, clipped, precise style owes something to her choice of isosyllabics over (say) Lawrentian free verse), but (being disengaged from the phonological facts of the language) they cannot offer the poet any precise control over rhythm.
Notional metres are usually the result of linguistically naïve attempts to borrow an inappropriate foreign model such as the isosyllabic Japanese haiku or the quantitative Latin hexameter, and their rules are consequently framed in terms of a misunderstood or fantasy version of the language (see, for example, Attridge on Elizabethan experiments in quantitative verse); actual metres, in order to be intelligible to a listener, must engage with the phonology of the language. Yet one reason why previous discussion of Wyatt's metre has tended to be inconclusive is that much of it has taken place in something of a theoretical vacuum, a world of unfalsifiable armchair theorising that hasn't seen the need to relate in any principled way the metrical descriptions it offers to the phonological properties of early modern English. One recent editor, for example, has proposed scansions like these, in which ictus of some kind is supposed to fall on both syllables of a word like "learneth" or "hideth":
Figure 1.
Shé that | me léarn|éth | to lóve | and | súf|fer
And thére | him híd|éth< | and nót | appéar|eth (Rebholz 51)
But if ictus (actual or notional) can occur on this weakest kind of syllable it can occur anywhere, and it will become impossible to write an unmetrical line. Needless to say, where unmetricality is impossible there can be no metre (even notional metre): such a system might be termed "vacuous". It seems unlikely that Wyatt would have remained permanently satisfied with a purely imaginary system of versification.
To anchor my discussion in the phonological realities of early modern English, therefore, I am using the linguistically-based method of scansion set forth in much greater detail in Groves (Strange Music), in which scanning a line involves mapping the prosodic shape of the verse as determined by phonology, syntax and pragmatics (upper tier of symbols) onto one of a number of pre-existing metrical patterns or templates (the lower tier); these templates are derived from the simple iambic pentameter matrix (five feet, [7] each consisting of a weak position followed by a Strong one) by transformations that permit the exchange of strong and weak positions within a foot (reversals) or between feet (swaps), provided that any such exchange is followed by a "normal" positioning (this prohibits, for example, double and final reversals). The prosodic tier records lexically stressed (A) and unstressed (O) syllables and syntactic breaks, and relations between adjacent syllables; an unstressed syllable adjacent to a fully stressed one within a syntactic constituent is "dominated" (represented as a lower-case "o") which effectively means it cannot carry an ictus (that is, be matched with an S-position); this is what makes nonsense of scansions such as #1. A dominated syllable that receives contextually-determined contrastive or focal accent, however, is enabled to carry ictus (its symbol is underlined and made uppercase). A syntactically subordinated stress (as with "long" in "the long love" said without contrastive or focal accent) is marked with an "a"; an a-syllable in an S-position is a "harsh mapping," rigorously avoided in neo-classical versification.
There are thus two kinds of deviation possible here: if the prosodic base can only be mapped onto a pattern that is irregularly formed, as in #2.1), with its three successive reversals, we call the verse "irregular"; if it cannot be mapped onto any derivable template, we term it "unmetrical." An irregular line will permit the full complement of five beats, but they will be distributed in unusual ways; an unmetrical decasyllable, on the other hand, will yield only four.[8] Wyatt's "The Longe Love" scans in this system as follows (irregular template sequences in blue; irregular mappings in green; unmetrical mappings in red):
Figure 2.
Critics who admire Wyatt's skill as a metrist have sometimes felt the need to show that he was writing modern pentameter, but the only way to scan some of these lines as modern pentameters is to distort the language, by shifting stress around and adding and subtracting syllables ad libitum. Rei Noguchi, for example, has attempted to show that Wyatt's decasyllables are metrical under the rules of Halle and Keyser's generative metric, an approach explicitly predicated on the axiom that poets do not (indeed, cannot) violate linguistic givens for the sake of metre. He can only do so, however, by assuming a wholesale violation of linguistic givens, arbitrarily re-assigning lexical stress metri causa (e.g. óffence, lovér 139). Jack Conner, on the other hand, seeks to build the necessary distortions into the language itself, hypothesising that from Chaucer to Wyatt all word-final consonants were given an automatic syllabic release—or, in other words, every word that ends in a consonant had a final schwa, as in Lancashire song. This is proposed as "a kind of miracle oil to ease the grinding gears of [fifteenth-century metrics], though the fact that every word-final consonant is involved puts us in the messy position of having rather too much oil" (Groves, "Water" 69): "The longë lovë thattë innë my thoughtë doethë harbarë". Conner's reassurance that this automatically generated syllable may nevertheless be suppressed at will offers the reader a large range of possible readings for any given line, though even with this latitude it cannot rescue lines like "with his hardines taketh displeasur" or "for goode is the liff ending faithfully" from #2. But the main reason to reject Conner's hypothesis is simply the complete absence of evidence to support it.
An older tradition of critics, such as Foxwell and Swallow, hypothesised that Wyatt was co-opting, on Chaucer's authority (but in a highly selective and arbitrary manner), some obsolete features of Middle English, such as variable stress on Romance loanwords and sounded final <e>: a line like "The longe love that in my thought doeth harbar" , for example, would scan if longe were treated ad hoc as an inflected ME weak adjective (as in "And madë revel al the longë night" (CT A2717), as would line 4 if baner were given Romance endstress.[9] These strategies, which postulate in Wyatt an implausible degree of philological sophistication (these features of Chaucer's versification were first identified by Tyrwhitt in the eighteenth century), would in any case recuperate only a proportion of the problematic lines: they would not, for example, permit scansions of the kind exemplified in #1 above. A. K. Foxwell enlarged the possibilities of linguistic distortion by proposing that Wyatt found authority for scansions like #1 in the corrupted text of Pynson's 1526 Chaucer; she saw him scanning the second line of Pynsons's version of the Canterbury Tales, for example, as "The dróught | of Márch | hath pér|cèd | the róte" (Study 39). To scan the line in this improbable fashion, however, Wyatt would have to have been rather firmly convinced that Chaucer intended to write pentameters: the snag is that nothing in the available printed texts would lead him to suppose that this was the case.
Indeed, the problem with all these Procrustean approaches is that they assume that the pentameter already existed as a model for Wyatt, and that his chief problem was to mould the language of his verse to fit it. But if we are to understand what Wyatt (or any poet) was doing we need to think historically, and to begin by considering the kinds of model of versification, both native and foreign, available to him. We also need to consider the three basic ways in which metrical systems are disseminated, which (as I have written elsewhere)
might be called the structuralist, the formalist and the impressionist. The most naive of these is the impressionist: such interpreters (who may not be aware that they are interpreting) simply attempt to reproduce in the target prosody what they think they hear in the source prosody. . . .The formalist, by contrast, attempts to reproduce what he or she knows or believes to be the formal features of the source line, with no regard for any phonological differences there may be between the target prosody and the source prosody: English haiku measured by syllables are formalist in this sense, as were those doomed Elizabethan attempts at quantitative hexameters. . . .The most sophisticated approach is the structuralist, which recognises both the abstract pattern of the source metre and the particularities of its phonological embodiment, and attempts to reproduce that pattern in the prosodic material of the target language, recognising that (as in the case of translation) all you can ever hope to achieve is a kind of approximation (Groves "Water" 55-6).
Within a linguistic community metres are perhaps most often propagated impressionistically, though (needless to say) the system doesn't necessarily work, as the home-made verses in the 'In Memoriam' columns of a newspaper will demonstrate. Where adequate metrical descriptions exist, as in the case of the Latin hexameter or French alexandrine, a purely formalist transmission is possible, but the English pentameter, lacking till recently an adequate formal description, has been for the most part transmitted impressionistically—and not always successfully, as victims of amateur Shakespeare can testify ("Now, fair HippoLY ta, our nuptial hour"). Occasionally writers have attempted to rely on inadequate formalist descriptions of the metre: Samuel Johnson "used once to be sadly plagued with a man who wrote verses, but who literally had no other notion of a verse, but that it consisted of ten syllables. Lay your knife and your fork, across your plate, was to him a verse" (Boswell 2.51).
If we turn now to the printed texts of Chaucer available to Wyatt, we see that they offer no consistent formalist evidence for decasyllabicity, and the only impressionistic assumption to be draw from them is that Chaucer's line was four-beat doggerel. This is because they are the result of a century-long process of corruption, an unintended scribal "reconstruction of Chaucer's metric, a (presumably) unconscious editing (based on an impressionistic reading of his metre) toward a rough four-beat norm" (Groves "Water" 58). The scribes who copied and re-copied Chaucer's lines failed to recognise its novel metre and presumably heard it as the familiar four-beat native line; consequently in transcribing it they inadvertently omitted function words and other minor syllables that enabled — and so would have been preserved in their short-term memory by — the pentameter pattern. In consequence, the verse that appears in Pynson (1526) or Thynne (1532) can only be consistently read as rough four-beat doggerel. Consider the opening lines of The Canterbury Tales as represented in Pynson's 1526 edition (text missing from Pynson but found in Hengwyrt and Ellesmere is represented in strike-out; pentameter-crippling omissions are in red):
Figure 3.
Whan that Aprill with his shoures sote
The drought of march hath perced [to] ze rote
And bathed euery veyne in suche lycour
Of whiche vertue engendred is the flour
Whan zepherus eke with his sote brethe
Enspyred hath / in euery holte and hethe
The tender croppes / and the yonge sonne
Hath in the Ram / [his half cours] halfe his cours yronne
And smale foules make[n] melodye
That slepen all [the] nyght with open eye
So pricketh hem nature in her corages
Than longen folke to gon on pylgrimages
And palmers [for] to seche[n] straunge strondes
To seruen halowes / couthe in sondry londes
And specially / fro euery shyres ende
Of Eng[e]land to Caunterbury thy wende
The holy blysfull martir for to seke
That them hath holpen / whan [that] they were seke. (CT 1-18, Pynson 1526)
Even if we permit the full range of template variation (including headlessness), two-thirds of the first 50 lines in Pynson's Chaucer are unmetrical — that is, they are incapable of being read in Early Modern English as pentameters, even as irregular ones — and only one need be so read.[10] An early modern reader of Pynson's Chaucer would have no reason to see it as anything but the kind of rough four-beat verse that Pope produced in his imitations of Chaucer: "|Whan that |Aprill with his |shoures |sote / The |drought of |march hath |perced z |rote".
But if Chaucer didn't provide Wyatt with a usable model, who did? George Wright, in his sensitive and scholarly analysis of Wyatt's metrical intentions, proposes Chaucer's younger contemporary Lydgate, who himself expressly modelled his practice on that of his 'maister Chaucer'.[11] As the pioneering work of Schick (lvi-lxiii) showed, however, Lydgate seems (understandably) to have misconstrued Chaucer's novel metre; he seems to have heard impressionistically the five beats of the pentameter, and the general alternation of beat and off-beat, and to have noticed the formal possibility of initial reversal, but to have superimposed on this (as Wright points out) the formalist requirement of the native metrical tradition for a heavy medial caesura and two independent cola, along with the relative indifference to syllable-count at the boundaries of the colon that typifies native metre. What we have, in other words, is a genuine hybrid: a partial impressionistic recognition of the new form combined with heavy formalist interference from familiar kinds of metrical behaviour. Thus the Lydgatian or 'split' pentameter consists of two independent cola separated by an intonational break (represented as a virgule in the MS tradition), and bounded by optionally filled positions (lowercase):
Figure 4.
x w S W S x | x S W S W S x
Only potentially beat-bearing syllables may occupy S-positions. Line thus vary in length from eight syllables to twelve or thirteen, as in these examples from the Siege of Thebes:
Figures 5 and 6.
Wright sees Wyatt as having developed this Lydgatian "jointed line," not merely by metrical variations — allowing, it seems, reversals anywhere except line-finally — but also prosodically, in that (like Foxwell) he seems willing to grant Wyatt extraordinary latitude in mapping the prosody onto the metrical template, with many violent mismatches between prosody and metrical template. Wright's scansion (as indicated by the diacritics over the vowels) seems to suggest that while #7 is to be interpreted as containing a metrical variation (a reversal), #8 and #9 require the arbitrary re-assignment of phonologically determined stress patterns metri causa:[12]
Figures 7, 8 and 9.
One problem with this analysis is that it predicts line-lengths of anything from 8 to 13 syllables, whereas Wyatt's heroic verse, as I have pointed out, is fairly consistently decasyllabic. But more importantly this proposed development of the jointed pentameter, with its metrical switches and prosodic plasticity, seems to me to produce a line in which the relation between prosodic material and metrical pattern has become so complex and elastic as virtually to defy communication of the metre. The first syllable in every colon might belong to one of two or three different positions, and the possibility of further dislocations within the lines makes it hard for the listener to get his or her bearings. Wright himself makes little attempt to explain how so complex a relation might be perceived by reader or auditor: indeed, as he says, "the knack of hearing Wyatt's rhythms vanished soon after his death" (149), and in any case "the trouble with Wyatt's system is that even someone who understands it may not see how individual lines are to be read" (150).
Of course, none of this proves that Wright is wrong; as Derek Attridge has shown, humanists who experimented with quantitative metrics were in general happy with the idea of a purely notional and incommunicable metre (though as humanists they would, of course, have had nothing but contempt for Lydgate's medieval rhythmus). But it is difficult to imagine that (as Puttenham says) "hauing trauailed into Italie, and there tasted the sweete and stately measures and stile of the Italian Poesie" that Wyatt would have been content merely to elaborate Lydgate's clumsy misconstruction of Chaucer's metric into an unwieldy Rube Goldberg machine, rather than seeking to become one of "the first reformers of our English / meetre . . . in all imitating very naturally and studiously [his] Maister Francis Petrarcha" (Puttenham 48-9, 50).
Wright's hypothesis, while it acknowledges Wyatt's exploration of the possibilities of the pentameter, puts him essentially at the end of a tradition, a victim of terminal over-elaboration, like the Irish Elk; what I want to suggest is that the mixture of familiarity and strangeness we find in Wyatt's rhythms is due to the fact that he is at the beginning of a tradition, our tradition: he laid the ground-plan for the metre of Shakespeare, Milton and Pope, even if the superstructures that they erected deviate from it in some important respects.
Our first question is: where did he get the form of his verse, if not from Chaucer or his followers? It is true that the pentameter proper did not die out completely between Chaucer and Wyatt; it was kept alive in the work of Scottish poets like Henryson and Dunbar, for instance, and occasional ones crop up, perhaps by accident, in the verse of Hawes, Barclay and Skelton. But it is doubtful that this minor tradition, lacking (as it seemed) the authority of Chaucer, would have had enough glamour as a model to attract someone who in other areas of poetics seemed to be attempting to "make it new". The obvious contemporary model here as in other matters was the Italian one, the endecasillabo. But what would imitating the endecasillabo mean, exactly? It is sometimes assumed that it is a form of iambic pentameter, but this is a misunderstanding: it is an isosyllabic line consisting of two unequal cola (either 4+6 [a minore] or 6+4 [a maiore]), each colon terminating in a lexical stress with a syntactic juncture at the next word-break (the characteristic eleventh syllable is just a consequence of the fact that Italian words are regularly paroxytonic). In the following example, stressed syllables are blue in even position and red in odd; metrically necessary (i.e. colon-terminal) stresses are underlined. It will be seen that most stresses fall either on even-numbered syllables or on the first, but that a very small minority of a minore lines, like the last of #10, have the pattern 4-7-10:
Figure 10.
It is possible that Wyatt experimented with a formalist adoption of the endecasillabo, but this is not in any case the form taken by his mature decasyllable: of the 14 lines of #2, for example, only six can be scanned as endecasillabi (2, 3, 7, 10, 13 and 14). The possibility remains that he misunderstood the formal requirements of the endecasillabo, seeing it as a purely isosyllabic metre with no stress requirements: this might (as I have suggested) explain some of his experiments but would not account for a poem like "They fle from me," where the lines vary from nine to twelve.
A further objection to a formalist adoption of the endecasillabo is that the result would be a merely notional metre in English (the reader is invited to experiment). So what are the options for producing intelligible metrical form? The phonology of English provides three basic possibilities:
Wyatt experimented with ictosyllabic verse in his poulter's measure, perhaps because of its promise of order as a relief from the apparent slovenliness and formlessness of late medieval versification. To judge from the merciful paucity of his efforts in that metre, however, it seems that he discovered long before his contemporaries the chief drawback of ictosyllabic verse, its monotony: not only is the metrical pattern rigid and unyielding in itself, but its absolute predictability destroys the possibility of tension between the metre and the prosody of a line. Prosodic variation becomes irrelevant because it cannot signal metrical variation, and the language of the verse is dragooned into a lifeless parade-ground regularity. The effect is to deaden our responsiveness to the language (obligatory pauses shown as solidi):
Figure 11.
So feble is the threde | that dothe the burden stay |
Of my pore lyff, In hevy plyght that fallyth in dekay, |
The great achievement of ictothetic verse (usually lumped with ictosyllabic verse in the traditional category "accentual-syllabic") is to escape this leaden-footed insistence, to exchange it for the dance and play of 'counterpoint'; the danger, which Wyatt does not seem immediately to have appreciated, is to allow so much free play that the relationship between prosodic forms and metrical matrix becomes unintelligible, and the remedy for this is to limit that play by rules. The principle seems to have been discovered twice in English, once by Chaucer (see Groves, "Water") and once by Wyatt, and in each case by the same means: an impressionistic misconstruction of the endecasillabo.
To explain: because most stressed syllables in the endecasillabo fall in even-numbered positions, and because Italian speech tends to preserve (other things being equal) roughly equal timing of syllables, whether stressed or not, the even-numbered syllables in the line will occur more or less at regular intervals, sounding to the English ear as though each alternate syllable carried a beat: "La don|na che 'l| mio cor| nel vi|so por|ta." Thus to an English ear (and I stress that this is an illusion) both main types of endecasillabo will tend to suggest, impressionistically, the ictosyllabic shape of a five-beat line with a variable number of stresses but a regulated number of syllables. Not infrequently, however, the first syllable of the endecasillabo is stressed, suggesting an initial reversal ("Tósto | che del | mio sta|to fu|ss[i] accorta"), and occasionally we find 4-7-10 lines with what Halle and Keyser would call a stress-maximum in odd position, as in "a me si vols[e] in sì novo calore," which in English would usually make the line unmetrical:
Figure 12.
Formally all this seems to indicate that ictus may occupy uneven positions, and while illustrating the most frequent variation — the initial reversal — doesn't immediately suggest any constraint on this possibility. Together this combination of impression and formal observation may have suggested to Wyatt (though not, of course, necessarily in any theorised way) the possibility of ictothetic verse, and specifically, of iambic pentameter.
The oddities we find in Wyatt's pentameter are for the most part of two kinds. The first is to do with the metrical rules (the rules that specify metrical patterns or templates), and springs from the fact that he didn't immediately see the need for certain kinds of constraints on transformations — in particular the constraint on double and final reversals, and the normal constraint on swapping, which requires a subordinated stress in the second position. Where reversals involve the exchange of positions within the foot, swaps exchange positions between feet; as in #13:
Figure 13.
Swaps in the modern pentameter can only occur with the first two syllables of ascending stress contours such as the mind's eye or to live thrall. It is tempting to say that Wyatt experimentally extended the domain of the swap to simple o-A sequences, as in #14, but this is to put the cart before the horse. The point is rather that as a pioneer of the form he seems not to have seen any need to restrict swaps to Õ-a:
Figure 14.
These constraints are part of the settled modern tradition of the pentameter, however, because its conventions are not completely arbitrary: if you're going to have a five-beat ictothetic decasyllabic line where the metrical pattern is communicable and obeys the phonological constraints of the language (these two requirements are, of course, related) the core rules of the modern tradition seem to be pretty close to the natural rules: "The constraints on optional transformation can be explained by the need to preserve metrical intelligibility; in effect, they ensure a re-affirmation of the matricial sequence after each modulation of it." (Groves Strange Music 108). These limits were discovered fairly soon (Wyatt may have begun the process himself) as the 'corrections' in Tottel's Miscellany show.
The second oddity of Wyatt's verse is more interesting: it is that he explored some of the potential of ictothetic verse — in particular, the possibilities of zero occupancy of metrical positions — that have been neglected in most of the subsequent tradition for what are essentially prescriptive reasons. Because the ictothetic line represents a sequence of abstract syllable-positions rather than a simple sequence of syllables, it can tolerate apparent gaps or absences. If you're counting syllables, then you've got to have syllables to count; if you're counting positions, a position need merely be implied — by an intonational break, say, or a slight dwelling on a pair of adjacent syllables.
The most common of these is the zero occupancy of w-positions, also known as catalexis. Although it was avoided as a solecism in the literary tradition before the twentieth century, has been exploited in more oral forms of pentameter such as Shakespeare's dramatic verse. Whether or not it is consciously registered, it is experienced by the reader or listener as an absence of something expected: thus headlessness creates a kind of initial abruptness that mirrors, for example, the suddenness of King Richard's volte-face in #15a, or the explosive anger or exasperation of the speakers in the second two:
Figure 15.
- ^ Stay, the King hath throwne his Warder downe. (R2 1.3.118)
- ^ Goe, take hence that Traytor from our sight, (2H6 2.3.102)
- ^ Gentlemen, importune me no farther (Shr. 1.1.48)
In contexts of staccato urgency, such as Horatio's desperate interrogation of the fleeing ghost, Albany's urgent command to rescue Lear and Cordelia or the dying Banquo's plea to his son, there may be more than one catalexis in a single line:
Figure 16.
a. ^ Stay: ^ Speake; ^ speake: I Charge thee, speake (Ham. 1.1.51)
b. Nay, send in time. / ^ Run, ^ run, O run (Lr. 5.3.247)
c. ^ Flye good Fleans, flye ^ flye ^ flye (Mac. 3.3.17)
But the problem with catalexis is that it looks like a mistake: there are plenty of classical precedents for coping with too many syllables in the line, but none (in Latin verse) for coping with too few. Traditionally, in any case, the ictothetic and ictosyllabic modes have not been distinguished, and the basis of pentameter from Gascoigne to Brooks and Warren has popularly been thought of as syllabic: pints of editorial ink have been spilt from the eighteenth century to the present to magic away Shakespeare's catalexes, those apparent ugly reminders of his lack of "art". The result of this prescriptive ban is that poets in the main tradition of the pentameter have avoided them for the same sort of reason that writers have avoided split infinitives and clause-final prepositions (even Shakespeare faithfully obeys the prescriptive rules in his non-dramatic verse). Only in the middle to late twentieth century, with its more cavalier attitude to prescriptiveness of all sorts, has the device again become available to poets (see Groves, "Larkin"). Wyatt uses catalexis to good effect in the scornful emphasis of "^let the old mule byte vpon the bridill" ("A spending Hand" 65) or his contemptuous refusal to "^ prayse Sir Topas for a noble tale / and scorne the storye that the knight ^tolde" ("My nowne Iohn poyntz" 50-1).[13]
Rests, or the zero-mapping of S-positions, are more problematic than catalexes and less common, because harder for the poet to signal,[14] and so much more dependent upon the reader's informed and intelligent co-operation, but they do occur in Shakespeare's dramatic verse. The most clearly signalled instances, pointed out by Fredson Bowers, occur in the love-duet in the fifth act of The Merchant of Venice. Six shared lines end in "In such a night"; of these the first two junctures are normal, the second are excessive (they have an extra offbeat or feminine caesura), and the third pair are defective — that is, they have a rest (represented as Ø):
Figure 17.
Where Cressed lay that night. / In such a night
And ranne dismayed away. / In such a night
To come againe to Carthage. / In such a night
That did renew old Eson. / In such a night
As farre as Belmont. Ø / In such a night
And nere a true one. Ø / In such a night (MV 5.1.6-20)
Philip Larkin in modern times has made creative use of the pentameter rest (Groves "Larkin" 718-19). One of the best examples of a rest in Wyatt's verse was pointed out by George Wright, who cites #18 from the rhythmically haunting poem "Vnstable dreme," and having shown that both possible scansions for it within his system produce phonological nonsense (either rétorníng or tó : lépe), suggests a third possibility that lies outside his system, a rest in the second foot:
Figure 18.
As he points out, the rest after retorning "expressively, but very unconventionally, conveys the heart-in-mouth excitement of the leap" (151).
In #19 I illustrate some of these attributes of Wyatt's proto-pentameter[15] by re-scanning the poem I began with, now revealed as metrical proto-pentameter throughout (catalexes and rests in green):
Figure 19.
It will be seen that this theory holds some potentially expressive implications for performance (the hesitant rests in lines 4 and 11 seem particularly appropriate) though (unlike the Procrustean theories outlined earlier) it doesn't require the performer to violate the phonological rules of English.
To sum up: Wyatt's proto-pentameter differs from the modern version in two major aspects: firstly, there is a greater variety of templates, because the metrical rules put no constraints on the number or sequence of reversals and swaps, and secondly, those templates are more accommodating because the mapping rules that relate prosodic strings to them are somewhat more relaxed, permitting both extended swaps and zero occupancy of positions (catalexes and rests). You might suppose that such a permissive metrical code would admit just about any possible circumdecasyllabic line into the fold, but there are lines even in Wyatt's verse that remain unaccounted for. In the two satires "My nowne John poyntz" (MJP) and "My mothers maydes" (MMM), for example, there are ten lines (4.6% of the total) that are not metrical proto-pentameters; of these, however, seven fit the pattern of the rare 4-7-10 endecasillabo described above and exemplified in the last line of #10 (stressed seventh underlined, preceding unmetrical o/S mapping in red):[16]
Figure 20.
the powre of them to whome fortune hath lent (MJP 8)
withowte regarde what dothe inwarde resorte (MJP 13)
I graunt somtyme that of glorye the fyer (MJP 14)
And he that sufferth offence withoute blame (MJP 70)
for mony poisen and traison at Rome, (MJP 98)
and in her langage as well as she cowd (MMM 41)
Madde if ye list to continue your sore (MMM 100)
It is hard to avoid the conclusion that Wyatt saw the 4-7-10 line with its distinctive anapaestic rhythm as a kind of occasional licensed departure from metricality (the most famous Italian example is perhaps from Dante's inscription over the gate of Hell: "per me si va nell'etterno dolore," Inf. 3.2).
Nevertheless, "the mystery of Wyatt" seems still to resist full and complete explanation: there remain three lines in MJP that cannot be explained by my theory, being neither proto-pentameters nor endecasillabi, all with (unmetrical) o/Smappings in position 4:
Figure 21.
With Venus and Baccus all their lyf long, (MJP 23)
With innocent blode to fede my sellff ffat, (MJP 35)
the letcher a lover and tirannye (MJP 74)
But an explanation that covers 98.6% of the lines in MJP and MMM is at least worth pursuing further.
I want to conclude by exploring some of the expressive possibilities of the proto-pentameter, as manifested in the "rhythmical subtlety" (Muir Life 245) of 'They fle from me," a subtlety neatly ironed out by whoever revised it for publication in Tottel's Miscellany (1557). Tottel's concern in his editing appears to be to turn a somewhat enigmatic poem into the literary equivalent of fast food, something he can sell as a straightforward sentimental instance of the 'lover's complaint' genre ("I vnkyndly so am serued"). As has often been noticed, the main focus of his emendations is metrical: at line 7, for example, he removes a couple of extrametrical syllables used by Wyatt to speed up a line about busy-ness, but for the most part he "repairs" catalexis, which occurs in more than a third of the lines. The (very approximate) isochrony or equal timing of beats of English speech means that catalexis tends to slow down the utterance of verse (the more syllables in a measure, the faster you need to talk to get them all in — and vice-versa). Tottel's redactor thus destroys that slow meditative movement, as of a man musing to himself, that is so much a feature of Wyatt's poem. Consider, for example, the three successive beats, all on long vowels, of " It was no dreme[:] I lay ^ brode ^ waking"(15), that suggest in their drawn-out cadences the speaker's astonishment at the strangeness of the memory: it is turned by the reviser into the jaunty rhythm of "It was no dreame: for I lay broade awaking". His revision of "Into a straunge ^ fasshion of forsaking" (17) to "Into a bitter fashion of forsaking" simultaneously removes a retarding catalexis and irons out a bit of mystery. To Tottel's speaker there is no mystery about what has happened, or how he should feel about it: he's been hard done by, and he wants the world to know (hence the little bow to the gallery at the end). Wyatt's speaker, by contrast, seems genuinely puzzled by his experience. The rest after "kyndely" [20) suggests his baffled groping for words: if she has indeed treated him "kindly" by leaving (and not just "unkindly," at Tottel reductively has it) then perhaps it has to do with her "kind," her nature evoked in the first stanza as "wild," simultaneously timid and predatory.
Perhaps one of the most interesting metrical emendations is in line 11. Wyatt's line is unmetrical — a 4-7-10 line — and Tottel duly emends. Some research by Reuven Tsur of Tel Aviv University, however, suggests that Wyatt may have intended something more subtle than an anapaestic ripple in line 11. He has investigated the kinds of recuperative performance-strategies adopted by experienced readers of verse when they encounter such unmetrical lines, and what they don't do is crudely distort given stress-patterns like a ballad-singer. If there were a syntactic break before position 7, such a line would be not be unmetrical (because the unstressed syllable in position 6 would no longer be fully dominated): an example would be Shakespeare's "Call in the Messengers | sent from the Dolphin" (H5 1.2.21), though even there you have to be sure to mark the syntactic break to avoid slipping into anapaests. Compare: #21 and the metrical revision #22, in which the coloured sequences (positions 6-9) have the same arrangement of stressed and unstressed syllables:
Figure 21.
Burnt after them to the bottomless deep (Paradise Lost 6.866)
Figure 22.
Burnt after Lucifer: bottomless deep (/ Open'd beneath …) < (my construct)
Tsur found that experienced readers of 4-7-10 lines try to fudge a metrical reading by forcing a break where none exists in the syntax: "Burnt after them to the | bottomless deep". But notice the effect of this strategy in line 11: the speaker who attempts to recuperate this line as pentameter is forced to introduce an artificial pause — a slight hesitation — before shoulders ("When her lose gowne from her | shoulders did fall" that is completely absent from Tottel's mechanical regularisation "When her loose gowne did from her shoulders fall"). The rhythm of Wyatt's line suggests a hesitancy, an awkwardness in the action itself that makes it seem more spontaneous and ingenuous, less like a smooth practiced courtly seduction, and this enhances our sense that the woman in the encounter with her "naked foot" is not of the court, is in some sense "wyld". Wyatt's speaker is puzzled, then, because he genuinely doesn't know how to evaluate his loss: if what he loves in her is her wildness, her freedom, then to possess her would destroy what he seeks to possess.
Figure 23.
Tottel's Version
THey flee from me, that somtime did me seke
With naked fote stalkyng within my chamber.
Once haue I seen them gentle, tame, and meke,
That now are wild, and do not once remember
That sometyme they haue put them selues in danger,
To take bread at my hand, and now they range,
Busily sekyng in continuall change.
Thanked be fortune, it hath bene otherwise
Twenty tymes better: but once especiall,
In thinne aray, after a pleasant gyse,
When her loose gowne did from her shoulders fall,
And she me caught in her armes long and small,
And therwithall, so swetely did me kysse,
And softly sayd: deare hart, how like you this?It was no dreame: for I lay broade awakyng.
But all is turnde now through my gentlenesse.
Into a bitter fashion of forsakyng:
And I haue leaue to go of her goodnesse,
And she also to vse newfanglenesse.
But, sins that I vnkyndly so am serued:
How like you this, what hath she now deserued?
All this explains, I think, why Wyatt's verse can be sometimes so strange, sometimes so familiar in its rhythms. He explored not only metrical possibilities (such as the double reversal and the extended swap) that were discarded as the tradition developed because they placed too much of a cognitive burden on the reader and listener (he may have begun this process of discarding himself, since some of his poems have few or none of them); he also explored some mapping-rule possibilities that remained unrealised in most of the tradition for what are essentially prescriptive reasons.
It might at first seem unlikely that Wyatt should (apparently) pass from what C. S. Lewis has called "first, his floundering, and then, after conversion, a painful regularity" to metrical experimentation: "in his poulter's he ticks out regular metre with the ruthless accuracy of a metronome. . . . It is immensely improbable a priori that the same man . . . should have gone on, beyond the regularity, to the subtlest departures from it" (Literature 225). But the question-begging progression Lewis imagines here is a result of his eliding the crucial distinction between ictosyllabic and ictothetic metre: what he calls "a painful regularity," a rigidly predictable sequence of beats and offbeats, is a feature —indeed, the defining condition —of ictosyllabic verse such as poulter's measure; both Wyatt's "floundering" and his "departures," on the other hand, are features of his ictothetic verse, and result from the fact that he didn't immediately discover the system of constraints on variation that define the limits of metrical intelligibility in the modern pentameter. "Regularity," or the post-Spenserian consensus on metrical and prosodic variation, had to be found by trial and error: it was the destination of the pentameter rather than its point of departure.
The purpose of Tottel's editing was, of course, strictly commercial: it was to turn a puzzling writerly poem into a saleable readerly commodity. The problem with Wyatt's poem is that it makes the reader work too hard (on every level): too much informed negotiation is required to realise it as pentameter, while Tottel's version falls (on the whole) effortlessly and naturally into its regular lines. For the pentameter to become useful —that is, to become widely intelligible — it too needed first to be Tottelized: it needed to be tamed and simplified into something plain and obvious, the drumming decasyllabon of Gascoigne and Gorbuduc, as a prelude to the emergence of the complex modern pentameter in the work of the great Elizabethans.
References
Footnotes
[1] Anonymous TLS reviewer of Tillyard's edition of the poems, 19 September 1929, quoted in Muir (Poems) xlvi.
[2] Clearly there is much variation in the pentameter over this period, but there is also a common core of acceptable variation for all poets, most clearly exemplified in the verse of Pope.
[3] Interestingly, about the same proportion (roughly 35%) of the lines in Zulfikar Ghose's isosyllabic decasyllables can be scanned as pentameter.
[4] At the conference at which this paper was presented I heard an eminent Spenserian refer repeatedly to "LUcifera" and "guYON," as in "And proud Lucifera men did her call" (FQ 1.4.12.1) and "Whom when the good Sir Guyon did behold" (FQ 2.1.42.1).
[5] <http://www.english.cam.ac.uk/vclass/class1/note6.htm>.
[6] Heine suggests implausibly that he mingles such lines with pentameters in his sonnets.
[7] The term "foot" is used to identify a wS sequence in the matrix, partly for economy of description (a line consists of five such units) and partly because exchanges within feet (reversals) seem significantly less constrained than exchanges between feet (swaps).
[8] A nine-syllabled line, of course, need have no more than three beats (o-A-o o-A-o o-A-o: "Deliver the letter tomorrow".)
[9] No doubt some recent French loan-words vacillated between Germanic forestress and Romance endstress, as does modern garage (borrowed just 100 years ago), but this would only affect a small proportion of the problematic lines.
[10] "To sér|uen há|lowes / cóuthe | in són|dry lóndes" (15); ironically, it is the textual corruption of ferne to seruen that produces this effect.
[11] Lydgate's Troy Book, ed. Henry Bergen, vol. 1 (London: Oxford University Press, 1906), 5.3521; 'Floure of Curtesye', 236.
[12] He suggests that "in an actual reading of such lines it is possible that stress was sometimes more evenly distributed between competing syllables" (145), but this is to misunderstand the nature of stress, which is a structural phonological feature, not a merely phonetic phenomenon.
[13] When catalexis is required between an a-A pair like "knight told," the effect is to force emphatic or contrastive accent on to the a-syllable: "and scorne the storye that the knight told."
[14] Unless, of course, they are built into the metrical pattern, as is the case in poulter's measure.
[15] The term is not entirely satisfactory, in that it suggests something crude and primitive: the main problem with Wyatt's metre is, on the contrary, that it is too sophisticated to be easily and readily intelligible to the modern reader.
[16] John Thompson, who cites the first three lines of #20 (pp.18-19), correctly sees the first two as less irregular than the third. The explanation for this is that whome and dothe are followed by minor syntactic boundaries which partly protect them from domination by the following stressed syllable, whereas of is not.
ISSN 1546-0401
